The present invention relates to a signal processing unit, particularly to processing of a signal outputted from an image capturing element (hereinafter, also referred to as an imaging device) that converts an optical image obtained by irradiating light on an object such as document to an electric signal.
There is available an image reader that converts an optical image obtained by irradiating light on an object such as document to an electric signal so as to obtain image data. This type of image reader is equipped with a signal processing unit that converts an optical image to an analog image signal by means of an imaging device such as CCD (charge coupled device) and then converts the analog image signal to a digital image signal so as to obtain image data.
In processing an analog signal, because waveform itself is a piece of data, there is a problem that the data cannot be transmitted if the waveform is distorted by incoming noise.
An analog signal outputted from an imaging device is converted into a digital image signal generally after being subjected to signal processing including the clamp processing for direct-current regeneration of the signal by cutting off a direct-current component and then adding a specified direct-current potential to stabilize a black level, sample-and-hold processing for sensing an image signal level, and signal amplification processing for matching the signal with a conversion range. However, because analog image signal is apt to be affected by noise, the linearity to the density of document is lost and so the image data to be processed cannot be free from distortion. Accordingly, it is difficult to obtain accurate image data.
In order to eliminate the above defect, there has been disclosed an image input system in which the processing for converting the output voltage of the density area to be obtained from a photoelectric converter to digital data is performed in two steps by the means of adjusting the output of the photoelectric converter and by the means of processing digital signal (refer to the Patent Document 1).
It is difficult to eliminate superposition of noise on the signal simply by means of analog signal processing. However, since digitalization involves a problem that higher voltage power supply is needed or that sufficient dynamic range cannot be realized, there has been disclosed a sampling processing unit, comprising a sample-and-hold circuit and an A/D conversion circuit installed after the sample-and-hold circuit, wherein the A/D conversion circuit is equipped with multiple latch circuits and either addition or subtraction is performed on the input signal and delayed output signal of the latch circuits (refer to the Patent Document 2).
[Patent Document 1]                Tokkaihei 10-233903 (Japanese Patent Application Laid-open Publication)        
[Patent Document 2]                Tokkai 2002-57581 (Japanese Patent Application Laid-open Publication)        
However, the technique disclosed in the Patent Documents 1 and 2 is how to digitalize the analog image signal effectively and accurately, and so it is based on effective utilization of the A/D conversion means. Besides, the signal amplification and black level adjustment employed in the analog signal processing are not digitalized yet. Accordingly, waveform is distorted due to the noise generated during signal processing and therefore image data may not possibly be transferred accurately, resulting in a problem that horizontal lines or uneven density is caused along the line progression direction and hence the image quality deteriorates when image data obtained from an imaging device is recorded on a recording medium.